DESCRIPTION The glutathione (GSH) transferases catalyze the addition of GSH to endogenous and xenobiotic molecules bearing electrophilic functional groups. As such they represent a major route for the metabolism and detoxification of alkylating agents. The enzymes exist as dimers of identical or closely related subunits. The three-dimensional structures of representatives of each of the five known gene classes (alpha, mu, pi, sigma and theta) of GSH transferases reveal that each subunit is divided into two domains (I and II) that interact in a head-to-tail fashion to form the dimer. There appear to be two basic subunit interface types, an alpha/mu/pi type and a sigma/theta type, the former having a specific hydrophobic ball-and-socket interaction and the latter being more hydrophilic. In spite of the extensive structural data little is known about the influence of the domain interactions on subunit or domain stability. The role of dimer interactions in the conformational stability and catalytic activity of GSH transferases remains largely unexplored. The objectives of this application are to define the thermodynamic stabilities and folding pathways of class sigma and class mu GSH transferases which have distinctly different dimer interfaces and to establish the contributions of specific intersubunit interactions identified by crystallography in determining the conformational stability of the homodimeric structures and the specificity of subunit-subunit associations. The objectives of the research plan will be realized through completion of the following four specific aims. (1) The thermodynamic stabilities and folding pathways of native class mu enzymes (i.e., the rat M1-1 and M2-2 isoenzymes) and a class sigma enzyme (S1-1) from squid will be determined in order to establish the influence of the very different dimer interfaces on the energetics of the subunit-subunit interactions. (2) The importance of a key hydrophobic interaction involving Phe56 in the class mu isoenzyme and a specific hydrophilic interaction involving Arg77(mu)/Arg68 (sigma) will be determined by site-specific mutagenesis. (3) Domain-domain interactions at the dimer interface will be investigated with chimeric homodimeric class mu enzymes in which domains I and II of the M1-1 and M2-2 isoenzymes have been switched. (4) Crystal structures of the mutant and chimeric enzymes will be determined. The research will be carried out in collaboration with the Protein Structure Function Research Programme in the Department of Biochemistry, University of Witwatersrand. Glutathione S-transferases exist as either catalytically inactive unfolded monomers or as catalytically active domain 1 : domain 2 head-to-tail dimers. The folding and unfolding is reversible and follows a two state mechanism interconverting folded catalytically active dimer with unfolded catalytically inactive monomer. The subunit:subunit interactions in the folded dimeric structure are thus important for both the stabilization of the association (dimerization) of the subunits and the stabilization of the tertiary structures of the folded subunits of the dimer. Preliminary data on a sigma class enzyme from squid suggests that there may be other detectable intermediates in the folding pathway, not just the unfolded monomer and folded dimer. The proposed research for the Dirr group will examine both folding and unfolding reactions. Functional probes to detect and quantitate the folded dimeric native structure are catalytic activity and binding of fluorescent ligands. Structural probes include a variety of spectroscopic, chemical, electrophoresis and proteolysis assays. The thermodynamics of folding/unfolding will be measured using differential scanning calorimetry.